Theory of Electrolytic Dissociation. 157 



note would be under 320 per second, and the second partial 

 of the highest would exceed thnt frequency. It was thus 

 hoped that near the beginning of the series the third partial 

 would be specially exaggerated by the belly^s resonance, and 

 near the end of the series the second partial might receive a 

 similar preference. But these anticipated effects, if realized 

 at all, are almost entirely masked by a number of other 

 resonances of which the belly seems capable. In fact, it 

 appears difficult to obtain a sought resonance from an upper 

 partial of any note, although such a resonance may be elicited 

 accidentally. 



Figures 21 to 39, forming the third and last column of the 

 Plate, show the curves obtained by a repetition of experi- 

 ments at pitches the same as those in the second column, the 

 string being again bowed at 90 cm. and observed at I4'8 cm. 

 But this time the leg of the optical lever was on the glass- 

 headed pin at 50 cm. The general agreement of photographs 

 forming a pair in any one line in the second and third columns 

 testifies to the satisfactory working of the method. The 

 results are, however, too complicated to be readily analysed 

 at a glance. The last figure has for the belly's motion a curve 

 whose components have frequencies 1 and 2 and relative 

 amplitudes of 1 and £. This would give for the sound-box a 

 resonance of the order 368 per second. 



University College, Nottingham, 

 April 7, 1905. 



XVII. A Correction. By Harky 0. Jones*. 



IN a recent paper in this Journal f on " Recent Investi- 

 gations bearing on the Theory of Electrolytic Dissocia- 

 tion/' L. Kahlenberg attempts to overthrow this most 

 important generalization in the field of Physical and In- 

 organic Chemistry. His method of reasoning on the basis 

 of his supposed facts will be passed over without comment. 

 It should be said, however, that in my opinion we have now 

 a very satisfactory explanation of the fact that the gas-laws 

 apparently do not hold for the osmotic pressures of concen- 

 trated solution, — there is combination between the solvent and 

 the dissolved substance, so that such solutions are really 

 much more concentrated than we should think them to be 

 from the amount of dissolved substance contained in them. 



The object of this note is to correct a statement made by 

 L. Kahlenberg, concerning the results of some of my own 



* Communicated, by the Author, 

 f Phil. Mag. (6) ix. p. 214 (1905). 



